Pancreatic cancers can be classified according to where in the pancreas the cancer is found or according to the type of cell the cancer has originated from. Pancreatic cancer can occur in the head, body or tail of the pancreas and symptoms can vary depending on where in the pancreas the tumor is located. 70-80% of pancreatic cancers occur in the head of the pancreas. The majority of cancers of the pancreas are exocrine in type, and greater than 90% of these exocrine pancreatic cancers are adenocarcinomas. Nearly all of these are ductal adenocarcinomas, wherein the cancer occurs in the cells lining the ducts of the pancreas. In addition, there are rarer types of exocrine pancreatic cancer, such as cystic tumors, cancer of the acinar cells and sarcomas. Cystic tumors are tumors that cause a cyst or fluid-filled sac in the pancreas. Sarcomas, a cancer of the connective tissue holding together the cells of the pancreas, are rare and most often occur in children.
In addition to exocrine cancers, endocrine cancers of the pancreas can occur. The endocrine cancers can be named by reference to the hormone that they produce, e.g., gastrinomas (which produce gastrin), insulinomas (which produce insulin), somatostatinomas (which produce somatostatin), VIPomas (which produce VIP) and glucagonomas (which produce glucagon). In addition, lymphomas of the pancreas can occur, although they are rare.
Pancreatic endocrine tumors (PET) may occur either sporadically or as part of multiple endocrine neoplasia type 1 (MEN1) syndrome (Kloppel, G., et al., Ann. N.Y. Acad. Sci. 1014:13-27 (2004)). These neoplasms are clinically classified as functioning (F-PET) or nonfunctioning (NF-PET), according to the presence of symptoms due to hormone hypersecretion. F-PETs are mainly represented by insulinomas. At diagnosis, metastatic disease is observed in only 10% of insulinomas but in up to 60% of NF-PETs, and most PET-related deaths are caused by liver metastasis (Kloppel, G., et al., Ann. N.Y. Acad. Sci. 1014:13-27 (2004)). The malignant potential among PETs varies greatly and cannot be predicted on the basis of histological appearance. In fact, the vast majority of PETs are well-differentiated endocrine tumors (WDET) and are defined as well-differentiated endocrine carcinomas (WDEC) only when invasion or metastases are identified (Kloppel, G., et al., Ann. N.Y. Acad. Sci. 1014:13-27 (2004)).
Pancreatic acinar cell carcinoma (PACC) is an extremely rare tumor type distinct from ductal adenocarcinoma and PET, although some overlap with PET is observed by both the expression of neuroendocrine markers in one third of the cases and the existence of mixed acinar-endocrine carcinomas (Ohike, N., et al., Virchows Arch. 445:231-35 (2004)). PACC is always malignant with a median survival of 18 months, which lies between that of pancreatic ductal adenocarcinoma and endocrine neoplasms (6 months and 40 months, respectively) (Holen, K. D., et al., J. Clin. Oncol. 20:4673-78 (2002)).
Little is known about the molecular pathogenesis of PETs (Kloppel, G., et al., Ann. N.Y. Acad. Sci. 1014:13-27 (2004)). Inactivation of MEN1 gene is the most frequent genetic event identified in sporadic PET, while mutations in genes typically involved in pancreatic adenocarcinoma are uncommon (Perren, A., et al., Ann. N.Y. Acad. Sci. 1014:199-208 (2004)). Even less is known regarding the molecular anomalies of PACC (Abraham, S. C., et al., Am. J. Pathol. 160:953-62 (2002)). No gene expression profile data is available for PACC and our understanding of gene expression changes that occur in PET is still at an initial phase (Hansel, D. E., et al., Clin. Cancer Res. 10:6152-58 (2004)). MicroRNAs are small (20-24 nucleotides) noncoding RNA gene products that serve critical roles in many biological processes, such as cell proliferation, apoptosis and developmental timing. To perform these functions, microRNAs negatively regulate the stability and/or translational efficiency of their target mRNAs (Ambros, V., Nature 431:350-55 (2004)). Currently, 313 unique mature human microRNAs are known, 223 of which have been experimentally verified in humans (www.microrna.sanger.ac.uk). Recent studies suggest that aberrant expression of particular miRNAs may be involved in human diseases, such as neurological disorders (Ishizuka, A., et al., Genes Dev. 16:2497-2508 (2002)) and cancer. In particular, misexpression of miR-16-1 and/or miR-15a has been found in human chronic lymphocytic leukemias (CLL) (Calin, G. A., et al., Proc. Natl. Acad. Sci. U.S.A. 99:15524-15529 (2002)). Aberrant expression of microRNAs has been linked to cancers and diagnostic/prognostic characteristics of specific cancer types can be distinguished based on their microRNA profiles (Caldas, C., and J. D. Brenton, Nature Med. 11:712-14 (2005); Croce, C. M., and G. A. Calin, Cell 122:6-7 (2005)). Functional studies also have linked aberrant microRNA expression to carcinogenesis (Chan, J. A., et al., Cancer Res. 65:6029-33 (2005); Cheng, A. M., et al., Nucleic Acids Res. 33:1290-97 (2005); He, L., et al., Nature 435:828-33 (2005); and Johnson, S. M., et al., Cell 120:635-47 (2005)).
The development and use of microarrays containing all known human microRNAs has permitted a simultaneous analysis of the expression of every miRNA in a sample (Liu, C. G., et al., Proc Natl. Acad. Sci. U.S.A. 101:9740-9744 (2004)). These microRNA microarrays have not only been used to confirm that miR-16-1 is deregulated in human CLL cells, but also to generate miRNA expression signatures that are associated with well-defined clinicopathological features of human CLL (Calin, G. A., et al., Proc. Natl. Acad. Sci. U.S.A. 101:1175-11760 (2004)).
Identification of microRNAs that are differentially-expressed in pancreatic cancer cells may help pinpoint specific miRNAs that are involved in pancreatic cancer (e.g., pancreatic endocrine tumors, acinar carcinomas). Furthermore, the identification of putative targets of these miRNAs may help to unravel their pathogenic role. The present invention provides novel methods and compositions for the diagnosis, prognosis and treatment of pancreatic cancer.